P-013 Laboratory medicine Poster

Accelerating Levey–Jennings review: a Python-built browser tool outperforms conventional quality control middleware

Audience Choice Award

To rate this poster, scan the QR code on the poster itself — voting only works on site. Every attendee has 3 stars and can give at most one per poster. A star is final and cannot be taken back.

Authors

J. Croxford presenting

Division of Chemical Pathology, University of Cape Town, Cape Town, South Africa. National Health Laboratory Service, Groote Schuur Hospital, Cape Town, South Africa

#0
J. Rusch

Division of Chemical Pathology, University of Cape Town, Cape Town, South Africa. National Health Laboratory Service, Groote Schuur Hospital, Cape Town, South Africa

#1
M. Korf

Division of Chemical Pathology, University of Cape Town, Cape Town, South Africa. National Health Laboratory Service, Groote Schuur Hospital, Cape Town, South Africa

#2

Abstract

Background Internal quality control (IQC) review is an important but time-intensive laboratory activity, requiring manual assessment of multiple analytes. Conventional middleware platforms are often hindered by inefficient navigation, fragmented workflows and suboptimal data presentation of Levey–Jennings (LJ) charts. These limitations contribute to variability in IQC interpretation. There is a need for user-centered digital tools that streamline IQC review while enabling consistent decision-making and improving user experience.

Objective To develop and evaluate a task-focused, browser-based IQC review tool by comparing review efficiency and user-reported usability with conventional middleware.

Method A code-driven browser-based IQC review tool was developed incorporating structured flagging workflows, interactive LJ visualisation, exportable review summaries and operating directly on raw IQC data without preprocessing. An early adaptive decision-support component was included to capture user decisions for future machine-informed refinement. In a prospective, within-user usability study, ten participants reviewed matched IQC datasets using both the new browser-based tool and conventional middleware. Review time was recorded for each platform. Usability was assessed using five Likert-scale domains covering ease of use, review efficiency, clarity of information layout, confidence in IQC decisions and platform preference.

Results Review time decreased from a median of 30 minutes (P25 22:10, P75 36:58) on conventional middleware to 6:24 (P25 5:35, P75 6:50) using the browser-based tool, representing a 79% reduction (p = 0.002). Usability ratings favoured the browser-based tool across all domains, with 100% of users rating it easy to use (vs 20% for conventional middleware), 90% reporting review efficiency (vs 20%), and 100% finding the layout clear (vs 40%). Preference for routine use was substantially higher (90% vs 10%), while confidence in IQC decisions remained high on both platforms (90% vs 80%).

Conclusion A custom-developed IQC review tool substantially reduced LJ chart review time and was strongly preferred over conventional middleware. These findings demonstrate the potential of targeted, locally developed laboratory informatics tools to improve workflow efficiency and usability, support more consistent IQC review and provide a platform with scope for future adaptive decision-support integration.